AI pin style devices but useful? with MLX

[turbosupply83]

Maybe I'm missing something obvious.

I'm mostly working with FLUX in SGLang on B200. The part I'm trying to understand is tail latency.

Would appreciate practical advice rather than theory.

For context, I'm not chasing leaderboard numbers. I mostly want the system to feel consistent day to day.

[flintriver50]

Maybe I'm missing something obvious.

That matches what I was seeing. I ended up in about the same place using llama.cpp.

[routeline23]

My first guess would be this: Start by verifying the driver/runtime combo, then test with a smaller prompt, then look at speculative decoding. Once you know the plain path works, then add the faster path back in piece by piece.